Caffeine Effect as Confounding Factor in Sepsis Diagnosis by Heart Rate Signal Processing

Apnea leads to respiratory arrest in premature infants, which decreases through the administration of caffeine by increasing the heart rate (HR). Nowadays, using electrocardiogram (ECG) signals, along with studying and comparing heart rate characteristics (HRC) in premature infants is considered as the most critical claim in the early detection of diseases, especially sepsis. Accordingly, this study investigated the effect of caffeine on HRC.

To this end, the raw ECG data of infants were collected from the Akbarabadi neonatal intensive care unit section and then processed in time and statistical domain. Next, the effect of caffeine on their HRC was investigated, and finally, HRC signals were analyzed fifteen minutes before and immediately after caffeine administration.

Before caffeine administration, the probability distribution of inter-beat (RR) intervals and the probability distribution of the R2 /R1 ratio were close to the normal distribution. According to previous studies, the irregularity of the signal in the diagram of the beat to beat RR interval indicates the infant health. However, these diagrams showed an abnormal distribution, and a specific uniformity was observed in the RR interval diagram after the administration of caffeine.

Based on the results of this study, changes in the infant’s HRC and its pattern should be identified after drug administration in order to evaluate the status of newborns, primarily through new methods of sepsis prediction in preterm infants. Eventually, the findings of this study enable clinicians to consider the drug effect as a confounding factor with a specific pattern in the signal without disconnecting diagnostic devices from infants for drug administration.